1898-1902. No. 33.] UPPER DEVONIAN FISH REMAINS. 11 



have had an arrangement similar to that demonstrated by CH. R. EASTMAN 1 

 in the case of a closely related American species. 



The sensory canals, however, are quite plain as usual. In describing 

 them 1 have chosen a nomenclature differing from EASTMAN'S, and I 

 prefer to follow that of BRIDGE and GOODRICH. On both sides of the 

 head 3 branches of the sensory canals are distinguishable, and these 

 join behind the eyes. (According to EASTMAN'S description in the hinder 

 central bone.) From this point two branches proceed; one, which first 

 runs outwards and then forwards to the outer side of the eyes, I have 

 named the infra-orbital canal; the other, which at first pro- 

 ceeds inwards until it meets the corresponding branch on the other 

 side, and then runs forward in an oblique direction inside the eye, I have 

 named the supra-orbital canal. The two supra-orbital canals 

 thus unite at an angle in a greatly abbreviated supra-orbital com- 

 missure. Finally the third branch proceeds backwards from the central 

 bone and may be called the post-orbital canal. 



In the form under consideration the post-orbital canals proceed almost 

 parallel as far as the central bone; the supra-orbital branches then form 

 a slight arc, with the convex side backwards in towards the supra- 

 orbital commissure; from that point they run forwards as usual between 

 the eyes in a divergent direction, forming an angle of about 40. In 

 front of the eyes they bend inwards and then converge slightly towards 

 the front edge of the head. This bend is seen on the left side. As 

 the front portion of the head is lacking, the length of these converging 

 branches cannot be determined with certainty. The form of the head, 

 however, goes to show that these branches may be reconstructed as in 

 fig. 2. The infra-orbital branches first run outwards obliquely in a 

 slight arc, the convex side of which is turned to the front, but quickly 

 bend forwards and on the left side may be traced as far as the part 

 outside the orbits. The course is more like an arc than an angle as 

 in most other species of this genus. 



One interesting feature is the hinder end of the p o s t - o r b i t a 1 

 canal. EASTMAN says that in the case of the carefully studied form 

 Macrop. rapheidolabis NORW. & Ow., they disappear beneath the sur- 

 face of the external occipital plate on either side close to the hinder 

 margin of the head shield, passing obliquely downward and inward 

 below the cranial roof, and in the living state presumably communicated 

 with the internal auditory sense organs. 



Mem. N. Y. State Museum, 10, 1907. Page 103. 



